entropy.corrected: Correction of the Finite Sample Size Effect
In MEET: MEET: Motif Elements Estimation Toolkit

Description Usage Arguments Details Author(s) References See Also Examples

Each training matrix is formed by a finite number of samples. The probability estimation error using the nucleotide frequency causes a bias on the uncertainty measurement

1	entropy.corrected(H, ErrorHX, HXmax)

`H`	Entropy
`ErrorHX`	Error Entropy
`HXmax`	Maximum Entropy

This function uses the results of correction.entropy and correctionaprox functions.

Joan Maynou <joan.maynou@upc.edu>

T. D. Schneider, G. Stormo, L. Gold, and A. Ehrenfeuch, The infor- mation content of binding sites on nucleotide sequences, J. Mol. Biol., vol. 188, pp. 415 Nov. 1986.

correctionaprox, correction.entropy

data(iicc)
data(RenyiOrder)
Factortrans<-iicc$Transcriptionfactor
HXmax<-iicc$HXmax
H<-iicc$Entropy[[1]]
correction<-correction.entropy(q,p=nrow(Factortrans),long=1,iicc)
ErrorHX<-slot(correction,"sderror")[nrow(Factortrans)]
entropy.corrected(H,ErrorHX,HXmax)